disable all caches and wait a hour for bootup and cpu-z ?

i didnt get more then ~1,8Ghz out of a Tualatin Celeron :>

matze79 wrote:

disable all caches and wait a hour for bootup and cpu-z ?

i didnt get more then ~1,8Ghz out of a Tualatin Celeron :>

But that would show up on CPU-z and the Verification page.
http://valid.canardpc.com/show_oc.php?id=420005
http://valid.canardpc.com/8csk21

Both of witch dont show it being disabled.

Maybe with PC166 SD-Ram and some Thing like Vapochill..

I can be 100% sure that this result was made without cache disabling.
http://hwbot.org/submission/3021815_terrarapt … z_s_56sec_632ms

matze79 wrote:

Maybe with PC166 SD-Ram and some Thing like Vapochill..

The P3 I linked is useing pc166 ram.

Insane results, that's for sure.

Liquid nitrogen. Very cool. 🤣 Probably an insane rig - would be cool to see pics of that second one with the insane FSB.

Other than that just a volt modded motherboard and a big bunch of chips to sift through, I guess.

I would love to see some benchmarks!
It should get close to low voltage / cache celeron Ms!

Arctic wrote:

I would love to see some benchmarks!
It should get close to low voltage / cache celeron Ms!

No way in hell that's happening. At these kinds of frequencies they are lucky if they can get a CPU-Z validation score. The score tyuper posted has a SuperPi result which you can use to compare!

I'd like to see these machines pass some vigorous Prime95 testing. 😉

mr_bigmouth_502 wrote:

I'd like to see these machines pass some vigorous Prime95 testing. 😉

Top clock speeds on HWBoot tend not to be stable.
But I bet some of the folks over there have gotten 2ghz-ish stable to the point that they can run p95 for a while.
If you ask me 1.6-1.8ghz would be the most that such a cpu would 100% stable along with the system. You can have a stable cpu, but ram and Northbridge too. it's harder than you think. I used to do comparative benchmarking in the past and I know my in and outs. I got out of it do to the cost and time. 1-2k on a New PC ever few months is hell. And a pc that thats only gets ued once or twice... 😵

mr_bigmouth_502 wrote:

I'd like to see these machines pass some vigorous Prime95 testing. 😉

That is not the point. You just need it to be stable enough to pass the benchmark you are currently running. It this case, open cpu-z and saving a verification file and screen shot.

This has been discussed since people started to overclocking and benchmarking, always poeple arguing about stability and 24/7 usage.

In my opinion overclocking a system to the point it has no practical use because it's unstable is nothing but a waste of time and you accomplish nothing by doing so.
.

PCBONEZ wrote:

In my opinion overclocking a system to the point it has no practical use because it's unstable is nothing but a waste of time and you accomplish nothing by doing so.
.

If you can get it stable it's more then worth it.
I had a x58 setup with a xeon running at 5ghz-ish once. Sure it was not practical but dang it was fast.

PCBONEZ wrote:

In my opinion overclocking a system to the point it has no practical use because it's unstable is nothing but a waste of time and you accomplish nothing by doing so.
.

This. If you can't get it stable with a high end air cooler, it is useless. Putting a tank of liquid nitrogen on top of your CPU can destroy it because it makes the CPU brittle. When you try to lift the nitrogen tank off, you can sometimes end up taking part of the CPU, socket, and circuit traces along with it.

If you go about thinking that way, yes it is pretty useless. But this kind of overclocking is just insane. It's meant to be a showcase and nothing more, it's directly related to the "Because I can" mentality that a lot of computer enthusiasts share. I find it incredibly impressive that a 1.4 Tualeron can even sustain 2.6GHz and boot into Windows.